To diagnose a canonical mismatch on a backlink’s referring page, compare the canonical URL in the page’s original HTML with the canonical URL found after JavaScript renders. If the two values differ, you’ve identified a mismatch to investigate. That comparison tells you what the page declares in those two states; it does not establish which URL Google selected.
Summary
- Record the referring page’s canonical URL in its original HTML and its rendered HTML. Keep both values separate from the referring page’s own URL.
- A difference between those two canonical values is the mismatch. Use the crawler’s HTML Canonical and Rendered HTML Canonical columns to see and document it.
- Treat the mismatch as a diagnostic signal, not an indexing result. If you can edit the page, align its JavaScript canonical with the original HTML declaration; handle the backlink’s destination as a separate URL question.
What a canonical mismatch means on a backlink’s referring page
A canonical mismatch is a difference between the canonical link in a page’s raw HTML and the canonical link in its rendered HTML after JavaScript runs. On a backlink’s referring page, that comparison concerns the referring page’s own canonical declaration—not the URL the backlink points to.
Keep three URLs conceptually separate: the referring page’s URL, the canonical URL declared by that page, and the backlink’s destination URL. The referring page is where the link appears. Its canonical declaration indicates a preferred URL for that page. The backlink’s destination is the URL the link leads to. These may be related, but they describe different things. A mismatch in the referring page’s canonical declarations does not, by itself, tell you whether the backlink points to a canonical or non-canonical destination.
The issue is specifically about whether the canonical found in original HTML agrees with the one found after JavaScript execution. Screaming Frog defines its Canonical Mismatch issue this way and distinguishes the raw HTML value from the rendered HTML value. If those values differ, the page has a raw-versus-rendered mismatch to investigate.
That finding is narrower than an indexing conclusion. It identifies conflicting declarations in two versions of the page’s HTML; it does not establish that the page or backlink is indexed, nor which URL a search engine selected as canonical. Those are separate questions.
The useful first decision is therefore straightforward: are you trying to diagnose the referring page’s own canonical declarations? If so, compare its raw and rendered values. Don’t treat the link’s destination as one of those values unless the page’s canonical declaration actually points there.
Find the canonical in the referring page’s original HTML
Start by recording the canonical URL in the referring page’s original HTML. Keep it distinct from the referring page’s URL so you can compare the declaration with the rendered value without mixing up what each URL represents.
In the crawler’s results, the raw value is identified as the HTML Canonical. Screaming Frog’s guidance describes this as the canonical URL found in the original HTML before JavaScript is processed. Note the value as shown, and retain the URL of the referring page alongside it as the page being diagnosed. The page URL identifies the document; the canonical value is what that document declares.
If you’re reviewing more than one page, record each page and its raw canonical value together. That makes it easier to compare the right pair of values later. For an individual page, a simple working note can use these labels:
- Referring page: the URL of the page containing the backlink.
- HTML Canonical: the canonical URL found before JavaScript execution.
Don’t substitute the backlink’s destination for the referring page’s URL or its canonical declaration. The backlink destination is a separate URL. This step records only what the crawler found in the original HTML; it doesn’t yet tell you whether JavaScript changes the declaration or which canonical a search engine will select.
If there is no raw canonical value in the crawler’s result, you can’t make a raw-versus-rendered comparison from that value alone. Avoid recording a guessed URL as the declaration. The mismatch diagnosis depends on comparing the raw value with the rendered value, so keep the evidence for each side clear.
Check the referring page’s rendered canonical after JavaScript
Next, render the referring page with JavaScript enabled and record the canonical URL found after execution. Compare this rendered value with the raw HTML Canonical you already noted.
The procedure in Screaming Frog’s SEO Spider guidance is to enable JavaScript rendering, then review the JavaScript tab and the Canonical Mismatch filter. The Rendered HTML Canonical column shows the canonical URL found after JavaScript has been processed. These labels give you the two values needed for the comparison: HTML Canonical before JavaScript, and Rendered HTML Canonical after it.
Keep the page URL with the rendered value, just as you did with the raw value. That helps prevent a common mix-up during review: comparing the raw canonical from one page with the rendered canonical from another. The two values should belong to the same referring page.
The result you want from this step is a recorded rendered canonical, not a conclusion about indexing. JavaScript execution lets the crawler inspect the rendered HTML for the canonical declaration. It doesn’t establish that Google chose that URL or that the referring page’s backlink has been indexed.
If you’re reviewing an issue report, use its page entry and the two canonical columns to identify the pair to compare. Keep the original and rendered values visible together in your notes. Once you have both, you can decide whether they match. This is the point at which the diagnosis becomes a direct comparison rather than an inference from a single canonical value.
Compare the referring page’s raw and rendered canonical URLs
Compare the HTML Canonical and Rendered HTML Canonical values for the same referring page. If they differ, record a raw-versus-rendered canonical mismatch; if they match, that comparison does not show this particular mismatch.
The difference can be described simply: the original HTML declares one canonical URL, while the rendered HTML after JavaScript execution contains another. Screaming Frog identifies different raw and rendered canonical links as the condition that triggers its issue. You don’t need to infer a mismatch from the backlink’s destination, a page’s URL, or a search result. The relevant evidence is the pair of canonical values from the same page.
For a clear record, preserve the two values and label which one came from which HTML state. For example, note “HTML Canonical” and “Rendered HTML Canonical” rather than writing only “canonical URL.” That distinction matters because the mismatch is the disagreement between those values. A note that keeps only one value can obscure what the comparison established.
If the values are different, the comparison establishes a declaration conflict to investigate. It does not explain why the declarations differ, prove that either URL is selected by Google, or establish the status of the backlink’s destination. Keep those conclusions outside the finding unless you have separate evidence for them.
If the values are the same, the raw-versus-rendered check has not found a mismatch for that page. That result answers this diagnostic question only. It isn’t a general assessment of the backlink or an assurance about which URL a search engine will treat as canonical.
How to Get Google to Index Your Website Faster→Interpret a canonical mismatch without assuming Google’s choice
A raw-versus-rendered mismatch shows that the page’s canonical declarations disagree; it does not show which URL Google selected. Keep the observed difference separate from any conclusion about Google’s canonical choice.
The distinction matters because a page’s declared canonical and a search engine’s selected canonical are not necessarily the same. Google Search Central says that even when a site explicitly designates a canonical page, Google might choose a different one. The guidance names content quality and technical signals as possible reasons for that choice. So, a crawler’s record of two different declarations is useful diagnostic evidence, but it is not proof that Google selected either one.
You can state the finding precisely: “The raw HTML canonical differs from the rendered HTML canonical.” That describes what the comparison found. Avoid extending it to “Google selected the raw URL” or “Google selected the rendered URL” unless you have separate evidence establishing that selection.
This also keeps the backlink question in proportion. A mismatch on the referring page is about that page’s canonical declarations. It is not, on its own, evidence that the backlink destination is indexed or not indexed, or that the backlink is being treated in a particular way. Those claims go beyond what the comparison establishes.
For the next decision, ask whether the referring page can be edited and whether its raw and rendered declarations should be aligned. If you can edit the page, the declarations can be reviewed and corrected. If you can’t, retain the mismatch as a diagnostic finding rather than treating it as a confirmed search-engine outcome.
Choose a repair for the referring page’s canonical mismatch
If you can edit the referring page, align its JavaScript canonical with the canonical specified in the original HTML, rather than having JavaScript change it to a different URL. Keep that repair separate from any decision about the URL the backlink points to.
Screaming Frog’s guidance reports Google’s advice not to use JavaScript to change a canonical URL to something different from the URL specified in the original HTML. It also warns that conflicting canonical link tags may lead to unexpected results. The practical repair direction is to make the canonical declarations consistent across the original and rendered HTML, then compare the two values again.
Before editing, be clear about which URL you intend to declare as canonical for the referring page. The page’s canonical tag and the backlink’s destination are separate. A change to the page’s canonical declaration should not be described as changing the backlink destination unless the link itself is also edited.
There is a related but distinct issue when a page links to a URL that has been designated non-canonical. Siteimprove describes resolving that issue by updating links on referring pages so they point directly to the canonical target URL. That is a link-target correction; it is not the same repair as aligning one page’s raw and rendered canonical declarations. Don’t collapse the two into a single diagnosis.
Editing the referring page requires access to edit its content, templates, or navigation links. If you have that access, decide which of the two issues you’re addressing: a mismatch between canonical declarations, or a link that points to a URL designated non-canonical. Make the corresponding change, then check the relevant values again. If your task is only to diagnose the raw-versus-rendered mismatch, document the differing values and avoid presenting that finding as proof of Google’s choice or of the backlink’s indexing status.
Content update: November 8, 2026.
How to Diagnose a URL That Was Indexed and Then Disappeared→